Abstract The field of medical x-ray imaging experienced a digital revolution in the early 2000s, with the spreading of digital radiography systems which are based on active matrix at panel imagers (AMFPI). Since then we have seen rapid development and clinical translation of large-area AMFPI based on amorphous silicon active matrix technology. However they have one major difficulty to overcome: producing x-ray quantum noise limited images at very low dose. Advanced applications of AMFPI, including tomosynthesis and cone-beam computed tomography have fueled the development of the next generation detectors, mainly in the ability of AMFPI to generate high quality images that are quantum noise limited and free from artifacts at low x-ray exposures and high frame rates. An alternative approach is to operate the detector in pulse mode for photon counting which provides higher dose efficiency through efficient noise rejection, quantum-noise limited performance, and optimal energy weighting. Photon counting systems are also not susceptible to memory artifacts. While the development of photon counting detector for computed tomography has been very promising, no commercial photon counting 2D sensors exists for radiography or mammography due to the simultaneous requirement for both high resolution and large area. Our hypothesis is that a direct-conversion amorphous selenium detector with unipolar time-differential (UTD) charge sensing and avalanche gain can yield a cost-effective and large-area photon counting imager with spectroscopic capabilities. The true impact of photon counting is to provide hyperspectral imaging (via multi-energy thresholding) to enable widespread application of contrast enhanced (CE) breast imaging with rapid acquisition and without motion artifacts (via simultaneous acquisition of high energy and low images during a single x-ray exposure). The objective of this proposal is therefore to fabricate and test a prototype photon counting imager using the proposed eld-Shaping multi-Well Avalanche Detector (SWAD). Conceptually the proposed SWAD imager employs four major components: (1) a photon counting chip, (2) multi-well pixel geometry, (3) amorphous selenium (a-Se) photoconductor deposited over the multi-well substrate for UTD charge sensing and avalanche gain, and finally (4) an image acquisition circuit board where the SWAD chip will be connected to. We expect to show that the proposed photon counting SWAD imager has quantum-noise-limited performance, high spectral sensitivity for energy weighting, and high frame-rates. Successful development of SWAD will lead to the first ever cost-effective and large-area photon counting detector for x-ray imaging. Although this is seemingly a high cost proposal, the technological innovation we develop will lead to the widespread clinical application of a more efficient and lower dose contrast-enhanced cancer screening system for mammography.